Univariate (conditional logistic regression) analysis of risks factors associated to CQ-treatment failure (recrudescence), Madagascar, 2006–2007.

¥Treatment failure was based on the World Health Organization 28-day drug efficacy test and monitoring and corrected by PCR genotyping as described in Materials ans Method section;*Chi-squared test;§Fisher exact test;†ANOVA test (or if P-value of the Bartlett's test for inequality of population variances was #OR: odds ratio;¶95% Confidence Interval;±Discordant alleles found in 16 isolates between day 0 (wild-type NYD) and day of recrudescence (13 single-mutants YYD and 3 double-mutants YFD);±±Discordant alleles found in 8 additional isolates between day 0 (wild-type NYD) and day of recrudescence (8 single-mutants NFD);1according to [18];2according to [19] and3according to [20].